One Crash Can Be Worse Than Seven — Why Severity Beats Count Every Time

One crash. That's what the SAFER snapshot showed for DOT-3567102. My operations manager was ready to approve them the same day — good equipment photos, responsive dispatch, reasonable rate. The other carrier in the queue, MC-4829103, had seven crashes and a 45th-percentile Crash Indicator BASIC score. He was cutting the seven-crash carrier and booking the clean one.

I asked him to read me the breakdown on that single crash. Fatal: 1, Injury: 2, Tow-away: 1. Eight months ago on I-76 in Pennsylvania. I ran a quick search. No press coverage, no settlement I could find. Wrongful death statute of limitations in Pennsylvania: two years.

We used the seven-crash carrier.

---

Crash count is one of the most common shortcuts I see in carrier vetting — not because brokers are lazy but because the number is right there, easy to reference, and it feels like it should mean something. "Three crashes" sounds worse than "one crash." But that intuition fails badly when you haven't looked at what those crashes actually were.

Here's the thing about how FMCSA builds the Crash Indicator BASIC score: it counts reportable crashes without weighting for severity. A tow-away parking-lot backing incident counts the same as a four-fatality highway collision. The percentile rank in SMS reflects frequency adjusted for miles traveled. It says nothing about whether someone died.

That's not a flaw in the methodology. The BASICs are designed to flag carriers who might need a compliance intervention based on patterns, not to quantify your litigation exposure as a broker. That's a different task, and it's yours to do.

The breakdown is buried one click away

The SAFER company snapshot shows crash data in a way most people skim past. In the "Crashes" section, there's a two-year rolling window with crashes broken into three columns: Fatal, Injury, and Tow-Away Only.

Those three columns tell you almost everything the count won't.

A carrier with seven tow-away crashes and zero fatalities, zero injuries, is accumulating low-speed equipment incidents — backing into dock bumpers, parking lot contact, tire failure on the shoulder requiring a service tow. Not great. Not the same risk profile as a carrier with one fatal and two injury.

A carrier with one fatal crash has caused at least one death on a public road with a commercial vehicle. That's a different conversation and a different file.

Under 49 CFR § 390.5, an "accident" for FMCSA reporting purposes is any incident involving a commercial motor vehicle that results in a fatality, a bodily injury requiring immediate medical treatment away from the scene, or disabling damage requiring a tow. Minor fender-benders where everyone drove away don't appear. The tow-away threshold is already a filter. What shows up on SAFER has already been screened — and a carrier still in the fatal column got through that filter.

The statute of limitations problem

Personal injury and wrongful death claims don't resolve overnight. Plaintiffs' attorneys take time to file. Discovery takes time. Cases often settle two or three years after the accident.

A carrier with a fatality crash eight months ago almost certainly has live exposure — either active litigation or a pre-suit demand still in the window. In most states, personal injury gives plaintiffs two to three years to file. Wrongful death varies but runs in the same range. A $2M to $5M wrongful death case in federal court is routine in trucking, and the plaintiff's firm has months or years left on the clock to name defendants.

If you book that carrier on a load today and they get involved in a second incident — even a minor one — you are now discoverable in the original case as part of the "what kind of carrier was this" narrative. You're not a party to the first crash. But your due diligence files and carrier approval decision are absolutely relevant when the question is whether anyone bothered to look at this carrier's history before putting them on the road.

Post-Montgomery, the federal preemption shield is gone. Before May 14, 2026, a broker in the 7th or 11th Circuit could argue that the FAAAA preempted state-law negligent selection claims. The Supreme Court in Montgomery v. Caribe Transport II, LLC shut that down unanimously. You can now be sued in state court for selecting a carrier you should have known was unsafe. A carrier with a fatal crash eight months ago and a case still likely open is a carrier whose history a plaintiff's firm will put in front of a jury.

The carrier I declined? Their one crash looks fine in the BASIC percentile — "Insufficient Data" because the inspector count was low. The Crash Indicator BASIC showed nothing concerning. On paper, excellent. In practice, they were carrying live exposure I didn't want in my file.

What MC-4829103's seven crashes actually looked like

The seven-crash carrier we used: Fatal: 0, Injury: 0, Tow-Away: 7.

I called their dispatcher and asked directly about a few of the incidents. Straightforward answers: two backing incidents at a receiver's dock with a notoriously tight entrance (they'd stopped using that receiver entirely), one tire failure on the shoulder requiring a service tow, a couple of low-speed dock contacts. No injuries, no litigation, no hesitation in how they answered.

Their 45th-percentile BASIC score looked worse than DOT-3567102 on the dashboard. Their actual crash profile was equipment incidents and bad dock geometry. Annoying, not catastrophic.

For the load I was covering — a 42,000-pound general freight move, no hazmat, standard lanes — the seven-tow carrier was the materially safer selection. And if that load had ever ended up in discovery, my file would have reflected a deliberate, documented choice between two carriers evaluated on severity, not just count.

The accident register most brokers never ask about

Under 49 CFR § 390.15, every motor carrier is required to maintain an accident register for a minimum of three years. The register must include: date and location of each accident, name of the driver, number of fatalities, number of injuries, whether hazmat was involved, and whether the driver or carrier received a citation.

This is a private document. FMCSA doesn't publish it. But you can ask for it.

There's nothing preventing you from including a provision in your carrier agreement that requires the carrier to make their accident register available on reasonable notice for any accident involving a load you tendered. Most carriers with clean histories don't hesitate. Carriers who hedge about it sometimes have a reason to hedge.

I've never had a carrier with zero fatalities and zero injuries in their SAFER data refuse the request. I've had two carriers with fatal crashes in their data get evasive when I asked follow-up questions about the register. One of them gave me inconsistent information about the accident date. Both got declined.

The register request isn't about catching liars — most carriers aren't lying. It's about establishing a documented diligence record that shows you went beyond the SAFER snapshot. And it's about filtering for the rare carrier that reacts strangely to a reasonable question.

Reading the breakdown like a decision tool

When I'm evaluating a carrier's crash history, here's the actual sequence:

Fatal count first. Any nonzero number triggers a manual review. Not an automatic decline, but an automatic manual review. When, where, any indication of settlement, any reason to think the case is still live. If I can't find evidence that it's closed, I treat it as open.

Injury count second. Multiple injury crashes in a 24-month window — especially with rising BASIC percentile — is a pattern that warrants a closer look at inspection history and vehicle maintenance BASICs. An injury crash is a real crash. Someone needed immediate treatment off-scene.

Tow-away count last. A high tow-away count tells me something about equipment quality and driver attention, but a tow-away is orders of magnitude lower severity than a fatality. Twelve tow-aways in 24 months on a 10-truck fleet is worth a conversation. It doesn't automatically mean what one fatal crash means.

The BASIC percentile gives you relative comparison against the national fleet. The breakdown tells you what you're actually dealing with.

How I document this

In DOTScreener, the SAFER crash breakdown pulls automatically — fatal, injury, and tow-away counts are visible in the carrier summary without having to navigate to the SAFER snapshot separately. When I'm building a carrier file for an approval decision, here's what goes in the record:

• Fatal crash count and approximate date(s) from the SAFER snapshot
• Whether I found any press coverage or public indication of settlement
• My rough assessment of statute of limitations exposure (I note that I'm not an attorney and this is not legal advice — but the notation exists)
• Injury crash count
• Tow-away count and whether there's a pattern worth noting
• Whether the crash history triggered a manual review call with the carrier
• A screenshot of the SAFER crash section as of the approval date

The screenshot matters more than most people realize. SAFER data changes. Crashes fall out of the 24-month window. A crash that was visible on the day you approved a carrier might not show up two years later when discovery starts. The contemporaneous snapshot is what establishes what you knew and when you knew it.

If there's a fatal crash in the window with no clear settlement indication, the file note reads: "Potential live exposure — pending litigation cannot be confirmed or ruled out. Approved for this load based on [reasoning]. Manual review completed." That takes 45 seconds. It's the kind of documented reasoning that makes a difference across a deposition table.

---

The carrier with one crash is not automatically cleaner than the carrier with seven. The carrier with seven tow-aways is not automatically more dangerous than the one with a fatality.

Count is a starting point. The breakdown is the actual signal. If you've been looking at the number without reading what's in the fatal column underneath it, you've been making decisions with the wrong data — and in the current legal environment, that matters.

— Mason Lavallet

Founder, DOTScreener.com